Menu Home

Five common writing mistakes new scientists make

As a professor, journal editor, reviewer, and mentor, I review a lot of writing. I come from a long tradition of mentors who focused on writing — during my PhD, I often heard stories of my grand advisor returning his students’ work covered in red line edits, and then I experienced the same when I turned in my first drafts. My own students now know that this is something they can expect from me: close reading and detailed feedback. It’s how I grew as a writer, myself! I still remember comments from individual reviewers about bad habits in my manuscripts (thanks, Reviewer #3!), and I hear myself passing on my advisor’s comments (in his voice, even!) as I edit my students’ work.

As I’ve found myself doing more and more editing lately, I’ve started noticing patterns — common issues that tend to disproportionately show up in student and early career work. I’m by no means saying I’m an authority on writing (though I have read Stephen Heard’s excellent book on the subject), and I’m still “early career” myself by most definitions. Writing is a craft that improves with practice, both in terms of reading and in being read. I firmly believe that we can always improve our writing, regardless of career stage, and that it’s devoting the time to being thoughtful about our own writing practice. With that in mind, I thought I’d share five common bad habits that undermine otherwise good writing.


Getting your paper back with virtual or literal red ink can be stressful, but editing and revision are an important part of improving as a writer. Image CC0.

1. The passive voice is being used. This is an understandable mistake, because not only are we taught to write passively in primary and secondary school science classes, but many of my colleagues still cling to the idea that solid, objective scientific writing must be in the passive voice. This not only makes your prose difficult to follow, it also tends to result in unnecessarily long sentences, which is especially bad when word counts are at a premium. Why is the active voice better? Because science is an active process, done by human beings: “I” and “we” statements are appropriate when describing action. “We deployed data loggers” is much nicer than “Data loggers were deployed,” but it’s also clearer — I, the reader, now know exactly who did that work. The active voice emphasizes the agent of the action — the doer. The passive voice emphasizing the object things are happening to. We are a storytelling species — we like a little drama! The active voice is more engaging to read by its very nature, which makes otherwise dry methods sections just a bit less tedious.

2. It is entirely likely that your prose is padded with extraneous, superfluous, or otherwise unnecessary additions; furthermore, the utilization of such redundant verbiage is arguably obfuscating your points (thus, in order to improve the clarity of your writing, it is highly recommended that you eschew such stylistic choices, including run-on sentences filled with fluff, padding, and filler). Perhaps a consequence of assigning papers with word counts, text padding is one of the most common issues with student writing, especially for those writing their first manuscripts. My example sentence has a few issues: 1) double (or triple)-dipping with adjectives when one would do, 2) it crams too much, abusing semi-colons and parentheses for nefarious purposes, and 3) it’s full of “junk” phrases that serve no purpose whatsoever. “In order to” is not necessary when “to” will do. “It is entirely likely that” can be replaced with a single word (likely): “Your prose is likely….” Words like “arguably” or “furthermore” or “thus” rarely do any heavy lifting in sentences, and are often implied anyway. “Highly” isn’t needed in front of “likely.”

In addition, the above example is rife with $100 words where $1 words would do. This is painful for me to say, as a Scrabble-playing word geek who actually enjoyed studying for the verbal section of the GRE, but: as much as I love a good “eschew,” “avoid” generally works just as well. “Utilize” is rarely more appropriate than “use.” Take opportunities to be creative, but not at the expense of clarity. People often assume the thesaurus will help them sound smarter, but instead leaves your reader thinking, “wow, this guy really loves his thesaurus.” Use fun words sparingly, and aim for clarity. Relentlessly go over your prose to remove junk. When you’re over your word count on an abstract or conclusion section, look to cut sentence padding first, before you start cutting your cool ideas.

3. Your prose is redundant. You keep making the same point over and over again. I have a suspicion that this bad habit comes in part from the tendency to recycle prose from grant applications, papers, or abstracts. I often find myself reading prose that has the same idea presented in multiple ways — sometimes word for word, from one paragraph to the next! Regardless of how this happens, redundancy highlights the importance of taking a break from your work, getting feedback from a friend, and even reading your writing out loud (you’re more likely to catch errors than if you skim the page reading to yourself). Redundancy is also usually a symptom of poor organization; a lack of structure can lead to circular writing, because you don’t know where you’ve come from and you don’t have a clear sense of where you’re going (see #5).

I don’t really have any brilliant tips for avoiding redundancy, except: don’t do it. If you’re revising your own work, you should be catching the places where this happens. Also, get comfortable with deleting your writing. If this is really hard for you, create a “holding bin” for cut words and paragraphs — even if you never end up coming back for them, it’s less painful than deleting outright. But I suspect that one of the reasons redundancy is so common is that it’s difficult to let go of the work that goes into writing — once you’ve got words on the page, you want to keep them, because to ditch them would be to erase all the labor of writing. This is not true. Every sentence you write is part of the process; you improve your manuscript not with your word count, but with your editing. Cutting words is part of the writing process, and sometimes it’s the most effective way to make your writing better. That doesn’t mean the initial writing was wasted — it’s all part of what got you to cleaner, stronger prose. Think of cutting words like a switchback on a mountain trail: it gets you to the top of the mountain, even though it feels like you’re going backwards for a bit.

4. This use of unclear antecedents is inappropriate. Antecedents are expressions that give meaning to a proform (usually a pronoun) in a sentence. For example, in the sentence “Jacquelyn ate a piece of cake, and it was delicious,” the antecedent to “it” is “cake.” You can read more about antecedents here. I often see sentences where the antecedent to the pronoun is unclear — this is especially common with the word “this.” “This” is often used to refer to whatever ideas the writer has in her mind when she’s writing — it’s the central argument being discussed. The problem is, to a reader, “this” has to have an antecedent (the thing that came before that the pronoun refers to). If you don’t set up the sentence properly, the reader doesn’t actually know what the “this” or the “it” (etc.) is referring to. For example, in the sentence “Jacquelyn ate half of a piece of cake and gave the other half to Jessica because she was really happy,” it’s unclear  whether “she” is referring to Jessica or Jacquelyn (presumably they were both happy, because cake, but you get the point).

In scientific writing, I most frequently see this problem (haha, see what I did there?) in the beginning of sentences and paragraphs. “This is a problem, because…” What’s the problem? The increase in temperature you cite in the previous sentence that motivates your study, or the population crash you measured as a response? When you find yourself saying “this,” check to make sure that “this” is clearly linked to an antecedent. Remember that your readers aren’t in your head, and the connections may not be intuitive. [Author’s note: after publishing this post, I found at least ten cases of unclear antecedents in this post that I edited for clarity. I probably missed some, too! Editing your own work is important, folks!]

5. I love cake. Your paragraph needs a topic sentence. Props to Stephen Heard for this one. Realizing the importance of topic sentences has helped me quickly diagnose structural problems in writing (both mine and my students’, who are now probably sick of hearing me talk about topic sentences). I often read sections of prose that lack a clear roadmap, with ideas that are all over the place, and that have too many ideas packed into one paragraph (this issue is often exacerbated by some of the other ideas above). A very quick test to see if you’ve got organization issues is to check your topic sentences: the first sentence of your paragraph tells your reader what the paragraph is about, and every sentence should serve that topic sentence in some way. As with any rule, you can be a little creative here, but checking your topic sentences are a great way to check for structural issues in your writing.

If you find that you frequently end up in a different place than where you started, you might consider outlining your writing. The outline is a roadmap; you want to make sure that you’ll take your reader through all the important stops on the way to your final destination. It’s worth thinking about the structure before you start writing a section (e.g., your Discussion). Otherwise, you end up taking more of a random walk than a straight line to your point, and it definitely shows. It also creates a lot more work for your editors, and thus for you, down the road. Think about organization early and often.

Bonus #1: You changed tenses mid-paragraph, and/or your methods are written in the present tense. Repeat after me: methods will be written in the past tense. You’re talking about things you did in the past, not things you are doing in the perpetual present. Relentlessly check your prose for the present tense, and also for changes in tense that happen mid-sentence or mid-paragraph.

Bonus #2: You’re not following SI conventions. According to the International System of Units (SI), there must be a space between a value and its unit. 5g should be 5 g, 100ml should be 100 ml, 30ºC should be 30 ºC, etc. Now you know!

I hope you found these tips helpful! I should say that I’m a scientist with a strong foundation in the Humanities, but I’m not an English major or a professional editor. Please use this advice as a starting point to think about your own writing, with the understanding that there are other important issues that I didn’t get into here. There will also be people who may disagree with any one of the points above, because writing norms are often subjective, vary by field, and are highly contentious (you will pry the Oxford comma from my cold, dead hands). My hope is that you approach your writing as a process, and remember that there are ways to improve at every career stage, including excellent books like Stephen Heard‘s, taking advantage of professional editing services, or joining a peer writing group. Ultimately, the best way to get better at writing is to do it, and to approach revisions and edits as an important part of getting better. Treat feedback like a gift, rather than a personal criticism. To para-quote Samuel Beckett: Ever written. Ever written badly. No matter. Write again. Write poorly again. Write better.*

*The original is one of my favorite quotes: “Ever tried. Ever failed. No Matter. Try again. Fail again. Fail better.”

Categories: Commentary Grad School Professional Development Tips & Tricks

Tagged as:

Jacquelyn Gill

47 replies

  1. As both a scientist and a professional copywriter, I agree with pretty much everything you said Jacquelyn. On a rough estimate I can remove up to one third of the words in an article I’m editing, without losing any content or meaning. Mainly because we are taught to write to a length, we don’t think about the weight of the words we employ. Or how to be concise. One exercise I often get people to do during a workshop is to write one paragraph of text. Then I ask them to write the associated tweet that goes with it.

    One thing I do notice with theses, is they tend to be completed chapter by chapter. So not only do you have repetition within but also across chapters. When proof reading a text, familiarity certainly breeds complacency, so a fresh pair of eyes is essential. Finally thank you for mentioning the SI structure – it is one of my top 5 annoyances and with engineers the most common culprits. Value space Unit – it is not that difficult!


  2. I do a relatively large amount of cleaning up of papers and proposals on which I’m co-author, as well as others’ papers and proposals. Excluding higher-level logical development and explanation, this excellent article describes the greater bulk of what I fix (over and over, year after year, it never seems to change). I have sent the link to this article to many concerned, and will use this link as a flail, so to speak.


  3. I understand that the active voice can be better. But what if I did 10 things? Wouldn’t it be tedious to be saying “I” for multiple sentences in a row in the methods.


  4. Hi Jacquelyn – cracking post! I spend much of my reviewing/AE effort muttering to myself around why writers cannot put a point across in plain, simple English. Althought Methods sections tend towards the specialist, I plead for authors to explain their central concepts as if they were trying to communicate to a non-specialist of their acquaintance. For me, the whole point of writing a paper is to communicate effectively and complex, jargon-filled prose rarely achieves that.
    My other pet peeve is the use of ‘significantly’ in a Results section: I can read the stats output, what I want to know is whether or not the author found something of biological (in my field) significance.


  5. Good advice to writers! I also enjoy Robert Day’s book ‘ How to write and publish a scientific paper’ (2nd edn 1983; ISI Press). It may be older but it’s full of no-nonsense advice. I love his version of the 10 commandments of good writing, which starts with ‘(1) Each pronoun should agree with their antecedent’ and goes on in similar vein (e.g. ‘(5) Don’t use no double negatives’ and ‘(7) When dangling, don’t use participles’ ) to finish with ‘(10) About sentence fragments.’ William Safire’s 10 commandments for general good English are similarly pithy. One of my personal dislikes in scientific writing is the over-use of emotive words such as ‘controversial’, ‘imperative’ and ‘enigmatic’…


  6. You ecapitulate Orwell “on politics and the Englush language” nicely. If you don’t know that piece, go find it and get the tattoo!


  7. Excellent. Especially re topic sentences.

    I would add: Use short sentences. They really improve readability. I learned this from the old SF Chronicle columnist, Stanton Delaplane.


  8. Love it. Although, while “avoiding the passive voice” is good advice in most circumstances, there are cases where the passive voice is the prefered construction. For example, when the author recently talked about an “object” which is about to be “done to,” it makes sense to start the sentence with that object rather than the random unfamiliar subject that is about to act on it. I love Steven Pinker’s advice on writing (passive voice is discussed at 23:00 in the following video


  9. Thanks for the post! I definitely need that kind of advice. But I wanted to ask other thing, which glacier is the one on your blog? It looks strikingly similar to one we are studying at the moment.


    1. I think it’s the Perito Moreno glacier, in Argentina! I liked it because it shows the glacier and the trees together, much like what I imagine my own study sites might have looked like 14,000 years ago.


      1. With extreme confidence I would say is Aguila/Dainelli glacier on the De Agsotini Fjord, Darwin Range Tierra del Fuego, Patagonia. A bit more south than Perito Moreno. I could send you a photo. And yes, those are Nothofagus Pumilio an old tree that has its roots in Antarctica.


  10. Great advice!
    I also like the Beckett quote a lot, but shouldn’t your para-quote be: Ever written. Ever written badly. No matter. Write again. Write poorly again. Write badly/poorly better.


    1. Perhaps if you’re taking a 1:1 substitution, but this is a good example of the point that, sometimes, following rules to the letter results in decreased clarity, so knowing when to bend them is helpful.


  11. Thank you, Jacquelyn. Useful.

    I would love to hear your view, if possible, about recommendations included in the section “Choosing between active and passive” in the book The Craft of Research (Booth, W. Colomb, G., Williams, J. 2008: 262-3)


  12. Nice posting, gibes with my experience. I would add “punctuation”. I see more and more odd uses of commas, for example. But the biggest problem seems to be with colons and semicolons, which few students seem to know how to use properly…


  13. I love this post! I think you make excellent points and suggestions. I just want to point out that there are journals/fields in which they don’t allow the use of pronouns like “we” or “I”. I disagree with this ban because I agree with you we should use more active voice in our methods and be more explicit with the fact we are people doing science, not just that “science is done”. However, some journals won’t allow that style of writing, so authors should be careful of that and check their journal rules.


  14. What about when people give a year to an early online release citation (e.g. Gill, 2018), when in fact the article is “in press” and may not appear in a published issue of the journal until the following year?


    1. This is tricky! I tend to cite those as “in press,” though some journals won’t allow it. And I feel like this is gotten so much mud year now that we have online versions of articles that may come out months or even a year before the print version. I definitely feel like we need some clear guidelines on this, but I suspect there’s a lot of variation in practice.


  15. I think the guide is excellent, and will share it with my graduate students (with whom I am currently waging a war on the passive voice and vagueness of explanation). The only thing I disagree with is the statement that methods must be written in the past tense. I wonder if it is discipline specific?

    I frequently write my methods in the present tense, and have not been asked to change them by reviewers or editors. (I work in geophysics, btw.) I can certainly see that if there was a definite period of observation, or an experiment that occurred in the past, it would be natural to describe it as such. But if you are describing data processing or modeling, which are ongoing or repeatable processes, the present tense is appropriate.

    I do use the past tense when referencing historical events (“the earthquake occurred on…’) or the older literature. Otherwise it is present tense all the way. I like to think of it as “we do it this way, and this is what it means”, rather than “we did these things”.


    1. But if the model was already run, and you’re showing results from that past run, it’s in the past, right? The paper isn’t updated every time your model is run after you submit it.


      1. Well, a deterministic model will return the same results in the future as it does now, and as it did in the past. So it is as much a present result as a past one. I think it sounds less convincing, perhaps, if it only showed it in the past!

        Liked by 1 person

  16. I deliberately use the passive voice in a lot of occasion, because “we” didn’t spent hours developing the method and processing samples in the lab, “we” didn’t decide on the best analyses and performed them, “we” didn’t plan and write the paper, most importantly, “we” didn’t work overtime to the point of nervous breakdown to stick to unrealistic deadlines and “we” didn’t work for free for months once funding ran out because the project timeline was again unrealistic. I have a long list of co-authors that did little, if anything, but they are on because, politics (this was specifically said to me by my supervisor when I asked him to back me up in removing authors that did not contribute significantly to the research). I have no problem using the active voice when it is true and earned (e.g. “we” definitely spent a lot of time in the field to collect samples!), but it pisses me off to share credit with people that do not deserve it, so use the passive voice when it is the case. Sorry, rant over.

    Otherwise, good tips!


    1. I totally understand the credit issue. And while I try to have a strong and clear set of guidelines about what constitutes coauthorship, I understand that not everyone is able to follow those for political reasons.

      I never interpret “we” as being all individuals, for what it’s worth! In this case, I would just aim to make the language as active as you can. There are ways to write methods that don’t suggest everyone did the thing, but which still read dynamically and straightforward.


  17. And using “increased” when they didn’t see it increase – eg females have an increased risk of xerostomia… when did it increase? When they became females? Follow that one back… It’s higher, not increased. And don’t get me started on “compared with” – used where “than” should be – eg scores were higher in males compared to females…


  18. In my admittedly-anecdotal experience, #2 and #3 are the most common big problems in graduate student writing. I have yet to encounter a grad student (my past grad student self very much included) who avoids #2 and #3.


  19. My PhD committee insisted that everything should always be in the present tense. So, I still use present tense unless it really doesn’t make sense to do so. The methods are telling the reader what is happening in this paper and so can be in the present tense.


    1. I disagree here, because the paper isn’t where the action is. The paper is literally describing something that happened in the past. Obviously go with what your committee wants, but be prepared for journals and reviewers to request otherwise.


      1. (this is slightly tangential, but) I like to draw a distinction between methods, which should be written in the past tense (“We _prepared_ the samples this way”) vs. analysis/conclusions, which should be present tense (“Our experiment _suggests_ that X causes Y”). I prefer conclusions in the present tense because they weren’t only true when we did the experiments – the conclusions should also be valid now!


  20. Excellent suggestions, and they never get old or become irrelevant. For anyone.

    Everyone has writing pet peeves. One of mine is the word “It” being used as a place holder (see point #4). Thanks for pinning it down.

    I often suggest that new writers reread their introductions without the first paragraph – did losing it harm the paper? What about the next paragraph? Often 2-5 paragraphs of many papers can be easily deleted without loss of clarity or meaning. In fact, clarity usually improves. Readers should know the point of the paper within the first 2 sentences.


  21. These are excellent rules for news writing, as well. Except for the SI units, I’ve made every one of these points at some point over the years talking to other newspaper reporters, and have had them made to me in critiques of my stories. I’m especially prone to “vague ‘this’ syndrome”- from now on I will imagine a woolly mammoth shaking its head in disappointment every time I use that word as the subject of a sentence.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: